GOSSIP: a method for fast and accurate global alignment of protein structures.

MOTIVATION: The database of known protein structures (PDB) is increasing rapidly. This results in a growing need for methods that can cope with the vast amount of structural data. To analyze the accumulating data, it is important to have a fast tool for identifying similar structures and clustering them by structural resemblance. Several excellent tools have been developed for the comparison of protein structures. These usually address the task of local structure alignment, an important yet computationally intensive problem due to its complexity. It is difficult to use such tools for comparing a large number of structures to each other at a reasonable time. RESULTS: Here we present GOSSIP, a novel method for a global all-against-all alignment of any set of protein structures. The method detects similarities between structures down to a certain cutoff (a parameter of the program), hence allowing it to detect similar structures at a much higher speed than local structure alignment methods. GOSSIP compares many structures in times which are several orders of magnitude faster than well-known available structure alignment servers, and it is also faster than a database scanning method. We evaluate GOSSIP both on a dataset of short structural fragments and on two large sequence-diverse structural benchmarks. Our conclusions are that for a threshold of 0.6 and above, the speed of GOSSIP is obtained with no compromise of the accuracy of the alignments or of the number of detected global similarities. AVAILABILITY: A server, as well as an executable for download, are available at http://bioinfo3d.cs.tau.ac.il/gossip/.

Mutations in LZTR1 drive human disease by dysregulating RAS ubiquitination.

The leucine zipper-like transcriptional regulator 1 (LZTR1) protein, an adaptor for cullin 3 (CUL3) ubiquitin ligase complex, is implicated in human disease, yet its mechanism of action remains unknown. We found that Lztr1 haploinsufficiency in mice recapitulates Noonan syndrome phenotypes, whereas LZTR1 loss in Schwann cells drives dedifferentiation and proliferation. By trapping LZTR1 complexes from intact mammalian cells, we identified the guanosine triphosphatase RAS as a substrate for the LZTR1-CUL3 complex. Ubiquitome analysis showed that loss of Lztr1 abrogated Ras ubiquitination at lysine-170. LZTR1-mediated ubiquitination inhibited RAS signaling by attenuating its association with the membrane. Disease-associated LZTR1 mutations disrupted either LZTR1-CUL3 complex formation or its interaction with RAS proteins. RAS regulation by LZTR1-mediated ubiquitination provides an explanation for the role of LZTR1 in human disease.

Protein functional epitopes: hot spots, dynamics and combinatorial libraries.

Recent studies increasingly point to the importance of structural flexibility and plasticity in proteins, highlighting the evolutionary advantage. There are an increasing number of cases in which given, presumably specific, binding sites have been shown to bind a range of ligands with different compositions and shapes. These studies have also revealed that evolution tends to find convergent solutions for stable intermolecular associations, largely via conservation of polar residues as hot spots of binding energy. On the other hand, the ability to bind multiple ligands at a given site is largely derived from hinge-based motions. The consideration of these two factors in functional epitopes allows more realism and robustness in the description of protein binding surfaces and, as such, in applications to mutants, modeled structures and design. Efficient multiple structure comparison and hinge-bending structure comparison tools enable the construction of combinatorial binding epitope libraries.

Interdomain interactions in hinge-bending transitions.

BACKGROUND: The mechanisms that allow or constrain protein movement have not been understood. Here we study interdomain interactions in proteins to investigate hinge-bending motions. RESULTS: We find a limited number of salt bridges and hydrogen bonds at the interdomain interface, in both the "closed" and the "open" conformations. Consistently, analysis of 222 salt bridges in an independently selected database indicates that most salt bridges form within rather than between independently folding hydrophobic units. Calculations show that these interdomain salt bridges either destabilize or only marginally stabilize the closed conformation in most proteins. In contrast, the nonpolar buried surface area between the moving parts can be extensive in the closed conformations. However, when the nonpolar buried surface area is large, we find that at the interdomain interface in the open conformation it may be as large or larger than in the closed conformation. Hence, the energetic penalty of opening the closed conformation is overcome. Consistently, a large nonpolar surface area buried in the closed interdomain interface accompanies limited opening of the domains, yielding a larger interface. CONCLUSIONS: Short-range electrostatic interactions are largely absent between moving domains. Interdomain nonpolar buried surface area may be large in the closed conformation, but it is largely offset by the area buried in the open conformation. In such cases the opening of the domains appears to be relatively small. This may allow prediction of the extent of domain opening. Such predictions may have implications for the shape and size of the binding pockets in drug/protein design.

Some guidelines for identification of recognition sequences: regulatory sequences frequently contain (T)GTG/CAC(A), TGA/TCA and (T)CTC/GAG(A).

Inspection of many proposed recognition signal sequences shows that TGTG/CACA, GAGA/TCTC or their triplet subsets, and TGA/TCA occur frequently. These repeated elements, conserved in recognition sequences from evolutionarily distant organisms, are likely to possess unique structural characteristics. Recurrence of these oligomers may aid in identification of further regulatory sequences in upstream or other regions. Another class of recognition sequences is GC-rich. At present there are only a few examples of this class. It is likely that these sequences function via a different mechanism.

(A)GGG(A), (A)CCC(A) and other potential 3' splice signals in primate nuclear pre-mRNA sequences.

Several 3' splice signals in nuclear precursor mRNAs have already been known for some time: the AG doublet on the left-hand side of the splice and a run of pyrimidines just upstream of it. More recently it has been noted that the YNYTRAY sequence (where Y is a pyrimidine, R a purine and N any base) is a branching-sequence participating in formation of a lariat structure. Keller and Noon have shown the existence of several putative consensus sequences at this site. In this work, extensive computations of the distributions of 256 quartets in all primate nuclear pre-mRNA intron sequences present in GenBank have been carried out. Several putative signals upstream and downstream of the 3' splice have been detected. These have been compared with the results obtained in analogous computations carried out on all nuclear pre-mRNA introns present in a combined eukaryotic file containing mammal, non-mammalian vertebrate, invertebrate and plant sequences. The distributions of the more interesting oligomers are shown here. Of particular interest are the putative (A)GGG(A) signal 60 nucleotides upstream of the 3' splice site and (A)CCC(A) 3-40 nucleotides downstream of it. A possible splicing model explaining these data and involving formation of alternative hairpin loop structures is proposed.

RNA folding is unaffected by the nonrandom degenerate codon choice.

The frequent suggestion that the nonrandom codon usage is explained by its forming more stable mRNAs is tested in 22 genes. Only the histones, globins, and the rat preproinsulin gene show a correlation between the preferred degenerate codons and the stability of the secondary structure of the their mRNAs. However, the examined members from the histone and globin gene families, both among the oldest, in evolutionary sense, eukaryotic genes, have a high GC content (approx. 56% compared to an average of 42% in all eukaryotes) which is reflected in their degenerate codon choice and thus in their more stable folding.

Asymmetry in the distributions of the four nucleotides at mRNA initiation and 3' termini sites: some geometrical implications.

All mammalian and non-mammalian vertebrate sequences in the database have been aligned by their transcription initiation sites, and separately, by their mRNA 3' termini. A simple analysis of the distribution of single nucleotide composition in the 1000 nucleotides around these sites yields surprising results. An asymmetric pattern in the behaviour of complementary nucleotides, i.e., C vs. G and A vs. T, in both these sites is observed. The four nucleotides also behave in an opposite manner around mRNA transcription initiation and 3' termini. This may suggest that these are signals affecting the intrinsic dynamics of the DNA structure and as such facilitate the first stages of the recognition process of the RNA polymerase in the neighbourhood of transcription initiation. These signals may also play a role in transcription termination rather than serve as signals for the mRNA cleavage/processing machinery.

Sequence signals in eukaryotic upstream regions.

In eukaryotes, as in prokaryotes, evidence is accumulating showing that transcription factors recognize and bind to certain promoter elements. Sequence and chromatin structure perturbation specify the transcription initiation site and govern its efficiency. Two oligomers have been implicated in these processes: the TATAAATA and CCAAT. In the present work, all mammalian, non-mammalian vertebrate and invertebrate sequences accumulated in the database have been aligned by their mRNA start positions and scanned for recurrences of the 32 complementary triplets. The more significant signals are summarized here. In particular, TAT/ATA recurs very frequently further upstream, at -275, in addition to the 'classical' -40 position. Downstream their level is very low. The CAAT box complementary triplet components are not among the more striking signals. Closer examination of the -275 region indicates that to a large extent the signal is due to both the ATAT and the TATA quartets. Comparison of the frequencies of these quartets at -275 with the CAAT quartet at -80 suggests that the former signal is twice the strength of the latter. The oligomers' distribution charts support notions that several components are involved in recognition, making the regulatory regions more robust and less sensitive to mutations.

Enhancer elements share local homologous twist-angle variations with a helical periodicity.

Several experiments have shown that some enhancers can be exchanged between different genomes. The transferred enhancers were functional (cf. Levinson, B., Khoury, G., Vande Woude, G. and Gruss, P. (1982) Nature 295, 568-572). This argues that these exchanged fragments are recognized as enhancers and possess some common characteristics which other sequences lack. Extensive comparisons of enhancers yielded only very limited nucleotide sequence homology, which appears to be insufficient for enhancer recognition. We suggest that the enhancers located and sequenced to date have recurring, periodic homologous twist-angle (tg) patterns. This helical periodicity and the symmetric nature of the repeating twist-angle features present a recurring spatial geometry. It also offers a possible explanation of the fact that inverted enhancers are still functional. Regions of large twist, roll or main-chain torsion angle delta deviations from regular B-DNA may facilitate enhancer recognition especially when distant from promoter elements. Tissue specificity may be encoded in additional sequence or structural features.

Distinct patterns in homooligomer tract sequence context in prokaryotic and eukaryotic DNA.

The distributions of the junction sequences of homooligomer tracts of various lengths have been examined in prokaryotic DNA sequences and compared with those of eukaryotes. The general trends in the nearest and next to nearest neighbors to the tracts are similar for both groups. In both prokaryotes and eukaryotes A/T runs are preferentially flanked on either the 5' or the 3' ends by A and/or T. G/C runs are preferentially flanked by G and/or C. There is discrimination against A/T runs flanked by G or C and G/C runs flanked by A or T. However, whereas the distribution of prokaryotic homooligomer tract junction sequences was quite homogeneous, large variations were observed in the 5-fold larger eukaryotic database, increasing in magnitude from tracts of length 2 to 3 to 4 base pairs long. Possible DNA conformational implications and in particular DNA curvature and packaging aspects of prokaryotes and eukaryotes are discussed.

Molecular dynamics simulations of a beta-hairpin fragment of protein G: balance between side-chain and backbone forces.

How is the native structure encoded in the amino acid sequence? For the traditional backbone centric view, the dominant forces are hydrogen bonds (backbone) and phi-psi propensity. The role of hydrophobicity is non-specific. For the side-chain centric view, the dominant force of protein folding is hydrophobicity. In order to understand the balance between backbone and side-chain forces, we have studied the contributions of three components of a beta-hairpin peptide: turn, backbone hydrogen bonding and side-chain interactions, of a 16-residue fragment of protein G. The peptide folds rapidly and cooperatively to a conformation with a defined secondary structure and a packed hydrophobic cluster of aromatic side-chains. Our strategy is to observe the structural stability of the beta-hairpin under systematic perturbations of the turn region, backbone hydrogen bonds and the hydrophobic core formed by the side-chains, respectively. In our molecular dynamics simulations, the peptides are solvated. with explicit water molecules, and an all-atom force field (CFF91) is used. Starting from the original peptide (G41EWTYDDATKTFTVTE56), we carried out the following MD simulations. (1) unfolding at 350 K; (2) forcing the distance between the C(alpha) atoms of ASP47 and LYS50 to be 8 A; (3) deleting two turn residues (Ala48 and Thr49) to form a beta-sheet complex of two short peptides, GEWTYDD and KTFTVTE; (4) four hydrophobic residues (W43, Y45, F52 and T53) are replaced by a glycine residue step-by-step; and (5) most importantly, four amide hydrogen atoms (T44, D46, T53, and T55, which are crucial for backbone hydrogen bonding), are substituted by fluorine atoms. The fluorination not only makes it impossible to form attractive hydrogen bonding between the two beta-hairpin strands, but also introduces a repulsive force between the two strands due to the negative charges on the fluorine and oxygen atoms. Throughout all simulations, we observe that backbone hydrogen bonds are very sensitive to the perturbations and are easily broken. In contrast, the hydrophobic core survives most perturbations. In the decisive test of fluorination, the fluorinated peptide remains folded under our simulation conditions (5 ns, 278 K). Hydrophobic interactions keep the peptide folded, even with a repulsive force between the beta-strands. Thus, our results strongly support a side-chain centric view for protein folding.

Salt bridge stability in monomeric proteins.

Here, we present the results of continuum electrostatic calculations on a dataset of 222 non-equivalent salt bridges derived from 36 non-homologous high-resolution monomeric protein crystal structures. Most of the salt bridges in our dataset are stabilizing, regardless of whether they are buried or exposed, isolated or networked, hydrogen bonded or non-hydrogen bonded. One-third of the salt bridges in our dataset are buried in the protein core, with the remainder exposed to the solvent. The difference in the dielectric properties of water versus the hydrophobic protein interior cost buried salt bridges large desolvation penalties. However, the electrostatic interactions both between the salt-bridging side-chains, and between the salt bridges and charges in their protein surroundings, are also stronger in the interior, due to the absence of solvent screening. Even large desolvation penalties for burying salt bridges are frequently more than compensated for, primarily by the electrostatic interactions between the salt-bridging side-chains. In networked salt bridges both types of electrostatic interactions, those between the salt-bridging side-chains, and those between the salt bridge and its protein environment, are of similar magnitudes. In particular, a major finding of this work is that salt bridge geometry is a critical factor in determining salt bridge stability. Salt bridges with favorable geometrical positioning of the interacting side-chain charged groups are likely to be stabilizing anywhere in the protein structure. We further find that most of the salt bridges are formed between residues that are relatively near each other in the sequence.

Homonyms, synonyms and mutations of the sequence/structure vocabulary.

Calladine (1982) proposed that steric repulsion between adjacent purines on opposite helix backbones accounts for the local variation seen in four helical parameters. By defining simple sum functions based on Calladine 's proposal, Dickerson (1983) has taken what he terms "the first step in establishing a sequence/structure vocabulary". In this letter we analyze the implications of the Calladine - Dickerson model with regard to " homonyms ", "synonyms" and mutations. Specifically, we (1) show that because of the number of adjacent helical positions affected by one transversion, two purine-pyrimidine sequences may be similar at the sequence level yet be very different structurally ( homonyms ); (2) list all sequences which, though different at the sequence level, share adjacent structural parameter values (synonyms); and (3) use two simple statistical measures to show that transversion mutations occurring between a purine and a pyrimidine (5'----3' on the same strand) are in general less disruptive of local helical structure than transversions occurring between a pyrimidine and purine. On the assumption that they are not inconsistent with experimental findings, we discuss the significance of these implications.

Protein binding versus protein folding: the role of hydrophilic bridges in protein associations.

The role of hydrophilic bridges between charged, or polar, atoms in protein associations has been examined from two perspectives. First, statistical analysis has been carried out on 21 data sets to determine the relationship between the binding free energy and the structure of the protein complexes. We find that the number of hydrophilic bridges across the binding interface shows a strong positive correlation with the free energy; second, the electrostatic contribution of salt bridges to binding has been assessed by a continuum electrostatics calculation. In contrast to protein folding, we find that salt bridges across the binding interface can significantly stabilize complexes in some cases. The different contributions of hydrophilic bridges to folding and to binding arise from the different environments to which the involved hydrophilic groups are exposed before and after the bridges are formed. These groups are more solvated in a denatured protein before folding than on the surface of the combining proteins before binding. After binding, they are buried in an environment whose residual composition can be much more hydrophilic than the one after folding. As a result, the desolvation cost of a hydrophilic pair is lower, and the favorable interactions between the hydrophilic pair and its surrounding residues are generally stronger in binding than in folding. These results complement our recent finding that while hydrophobic effect in protein-protein interfaces is significant, it is not as strong as that observed in the interior of monomers. Taken together, these studies suggest that while the types of forces in protein-protein interaction and in protein folding are similar, their relative contributions differ. Hence, association of protein monomers which do not undergo significant conformational change upon binding differs from protein folding, implying that conclusions (e.g. statistics, energetics) drawn from investigating folding may not apply directly to binding, and vice versa.

A study of four-helix bundles: investigating protein folding via similar architectural motifs in protein cores and in subunit interfaces.

Four-helix bundles are identified and characterized in the subunit interfaces of protein multimers. We find that this motif occurs as often in the interfaces as in the protein monomers. Common and different characteristics demonstrated by the bundles in the two environments suggest the possible stabilization mechanisms of the bundles via cooperative helical twist, dipole alignment and interhelical connections. Nucleation of parallel helix pairs may be a favourable pathway before the pairs couple into bundles during folding. Certain properties found chaotic in the interface four-helix bundles indicate that either the subunit association is far from the global minimum conformation, or that the footprints of the folding pathway are recorded in these properties.

DNAase I hypersensitive sites may be correlated with genomic regions of large structural variation.

Helical-twist, roll and torsion-angle variations calculated by the Calladine (1982)-Dickerson (1983) rules were scanned along several nucleotide sequences for which DNAase I cleavage data are available. It has been shown that for short synthetic oligomers DNAase I cuts preferentially at positions of high helical twist (Dickerson & Drew, 1981; Lomonossoff et al., 1981). Our calculations indicate that DNAase I sensitive and hypersensitive sites in chromatin are correlated with regions of successive, large, helical-twist angle variations from regular B-DNA. In many cases these regions exhibit large variations in base-pair roll and backbone torsion angles as well. It has been suggested that DNAase I cuts in the vicinity of cruciforms. However, it was recently demonstrated by Courey & Wang (1983) and Gellert et al. (1983) that such cruciform formation in a negatively supercoiled DNA is kinetically forbidden under physiological conditions. We thus propose that clustering of large twist-angle (and/or roll and backbone torsion angle) variations may be among the conformational features recognized by the enzyme. Specific cuts can then preferentially occur at base-pair steps with high helical twists.

A geometry-based suite of molecular docking processes.

We have developed a geometry-based suite of processes for molecular docking. The suite consists of a molecular surface representation, a docking algorithm, and a surface inter-penetration and contact filter. The surface representation is composed of a sparse set of critical points (with their associated normals) positioned at the face centers of the molecular surface, providing a concise yet representative set. The docking algorithm is based on the Geometric Hashing technique, which indexes the critical points with their normals in a transformation invariant fashion preserving the multi-element geometric constraints. The inter-penetration and surface contact filter features a three-layer scoring system, through which docked models with high contact area and low clashes are funneled. This suite of processes enables a pipelined operation of molecular docking with high efficacy. Accurate and fast docking has been achieved with a rich collection of complexes and unbound molecules, including protein-protein and protein-small molecule associations. An energy evaluation routine assesses the intermolecular interactions of the funneled models obtained from the docking of the bound molecules by pairwise van der Waals and Coulombic potentials. Applications of this routine demonstrate the goodness of the high scoring, geometrically docked conformations of the bound crystal complexes.

Molecular surface complementarity at protein-protein interfaces: the critical role played by surface normals at well placed, sparse, points in docking.

Rigid-body docking of two molecules involves matching of their surfaces. A successful docking methodology considers two key issues: molecular surface representation, and matching. While approaches to the problem differ, they all employ certain surface geometric features. While surface normals are routinely created with molecular surfaces, their employment has surprisingly been almost completely overlooked. Here we show how the normals to the surface, at specific, well placed points, can play a critical role in molecular docking. If the points for which the normals are calculated represent faithfully and accurately the molecular surfaces, the normals can substantially ameliorate the efficiency of the docking in a number of ways. The normals can drastically reduce the combinatorial complexity of the receptor-ligand docking. Furthermore, they can serve as a powerful filter in screening for quality docked conformations. Below we show how deploying such a straight forward device, which is easy to calculate, large protein-protein molecules are docked with unparalleled short times and with a manageable number of potential solutions. Considering the facts that here we dock (1) two large protein molecules, including several large immunoglobulin-lysozyme complexes; (2) that we use the entire molecular surfaces, without a predefinition of the active sites, or of the epitopes, of neither the ligand nor the receptor; that (3) the docking is completely automated, without any labelling, or pre-specification, of the input structural database, and (4) with a single set of parameters, without any further tuning whatsoever, such results are highly desirable. This approach is specifically geared towards matching of the surfaces of large protein molecules and is not applicable to small molecule drugs.

A dataset of protein-protein interfaces generated with a sequence-order-independent comparison technique.

While there are a number of structurally non-redundant datasets of protein monomers, there is none of protein-protein interfaces. Yet, the availability of such a dataset is expected to provide an added insight into a number of investigations. First and foremost among these is analyzing the interfaces to obtain their prevailing architectures, the forces that account for the protein-protein associations and their packing considerations. Their comparisons with those of the monomers are likely to shed additional light on protein-protein recognition on the one hand and on the folding of the polypeptide chain on the other. Docking simulations are also expected to benefit from the existence of such a dataset. A major stumbling block to the generation of a dataset of interfaces has been that the interface is composed of at least two chains. Furthermore, in the interfaces, each of the chains might be represented by non-contiguous pieces. Their order in the interfaces being compared might be different as well. This discontinuity stems from the definition of an interface. An interface consists of interacting residues between the chains, and those that are in their vicinity in the supporting scaffold, within a certain distance threshold. This necessarily yields unordered fragments, as well as isolated residues. Our novel, efficient, sequence-order-independent structural comparison technique is ideally suited to handle the task of the generation of a library of structurally non-redundant protein-protein interfaces. As it is computer-vision based, it views atoms as collections of points in space, disregarding their chain connectivity. In this work, 351 interface-families are created. Comparisons of the interfaces, and separately, of the chains which contribute to them, yield some interesting cases. In one of the cases, while two interfaces are similar, the structure of only one of the two chains is similar between the two complexes. The structure of the second chain of the first complex differs from that of the second chain of the second complex. Here the structure of the cleft in the first chain dictates the specific binding interactions. In another case, while the interfaces in the two complexes are similar, both chains composing them differ between the complexes. Lastly, the chains composing the complexes are similar, but the interfaces are dissimilar, providing a set of data for investigations of the favorable orientations of protein-protein associations.